Most people who go out to buy a new camera are shopping for megapixels. The more MP the better, right? Well, yes . . . but there are more important factors than the total number of pixels the camera can capture. And there is more than one way to measure pixels.

Ok, a quick diversion into color space. For all intents and purposes, electronic image formats see pixels in three colors – red, green, and blue (I”m simplifying here, so if you really want the details I will be talking plenty about color space in other posts). Your eye also uses three colors to see, so it makes sense that we replicated that system to reproduce images electronically. Anyway, if you see “purple” it”s actually being produced on your screen by:

  • red pixel: on
  • blue pixel: on
  • green pixel: off

So every “color” requires three pixels. This applies both to the way the camera captures the image (the camera “sees” red pixels, blue pixels, and green pixels) as well as the way your monitor displays the image (with little red, blue, and green lights). But when you save this image, “purple” only takes up one “pixel” in the file, not three. So some manufacturers count the number of *file* pixels that the camera can caputer, and other manufacturers count the number of *sensor* pixels (which is 3x the number of file pixels for the same camera).

But wait, it gets better. Just because the camera”s sensor *records* 12MP of data doesn”t mean that it recorded each pixel acurately! In fact, the amount of noise that can occur in a high density sensor sometimes outweighs the benefit of having more pixels. So why do companies makes cameras with useless pixels? Because most people buy based on megapixels. Oh, and did I mention that 80% of computer monitors (at the time of this writing) are configured to display . . .0.7 megapixels. And a large portion of those monitors are hard capped at 0.7 megapixels. That means you are only using 10% of the pixels in your high resolution image!

Don”t get me wrong, there are lots of reasons to want more pixels than you see. But for purposes of demonstration, let”s take a look at what happens when you have lots of “ok” pixels versus a few “good” pixels. For this example “good” pixels acurately represent the original scene, whereas “ok” pixels have noise in them – characteristic of what one might find in a cheap camera. Think of it as static in the camera”s sensor. Good cameras are static free, but it”s difficult (and expensive) to eliminate the static.

The two images below are identical, shot with a 10 megapixel (good) camera. The image on the top was initially downsampled to just 1 megapixel (throwing out 90% of the original pixels), but the image “quality” was maintained. The image on the right was left at 10 megapixels, but 10% of the pixels were given random static noise. Both images were then resized to a typical monitor viewing size. So which do you prefer, more pixels, or better quality pixels?

Small clean sensor (top) vs big noisy sensor (bottom)

Small clean sensor (top) vs big noisy sensor (bottom)